Fiber-to-the-home (FTTH) has been regarded as the ultimate form of broadband access offering very high bandwidth to end-users. Today's FTTH systems are mostly offered through point-to-multipoint time division multiplexed (TDM) passive optical networks (PONs) using a 1:N power splitter (e.g., Ethernet-PON, Gigabit-PON, and 10G versions of these systems) at a remote node (RN) in the field to share a common transceiver at the central office (CO), or through point-to-point (pt-2-pt) optical Ethernets with individual home-run fibers.
The upstream and downstream signals of a TDM-PON are transmitted using different optical wavelengths (usually 1310 nm for upstream transmission and 1490 nm for downstream transmission). The TDM-PON media access controller (MAC) within the CO schedules the transmission between the CO transceiver (TRX) and the end users by assigning appropriate time slots to each end user. A TDM-PON provides beneficial savings in the number of trunk fibers (between RN and CO) and optical transceiver counts at the CO while saving patch panel space to terminate fibers, but does not scale well with bandwidth growth. The bandwidth per household is often oversubscribed as the bandwidth per optical line terminal (OLT) TRX at the CO is shared among all optical network units (ONUs) connected to the given OLT TRX. To support Gb/s per user transmission speeds on a TDM-PON can require >10 Gb/s transceivers at each ONU. Thus, high-speed transmissions can be both technologically challenging and expensive.
Pt-2-pt optical networks provide very high bandwidths to end users, but do not scale well with optical fiber termination at the CO and fiber counts. Rather, pt-2-pt optical networks result in large numbers of trunk lines and transceivers and fiber terminations in the CO. This usually results in greater space requirements, higher power, and increased capital expenses.
A wavelength division multiplexed (WDM) PON is another approach, which provides the benefit of fiber consolidation and pt-2-pt virtual links to end-users by assigning separate wavelengths between the CO and individual users. It can offer the benefits of both pt-2-multipoint TDM-PON and pt-2-pt architectures. Traditional WDM-PON systems use a wavelength demultiplexer (as opposed to the power splitter used in TDM-PON) at the RN in the field to distribute a separate wavelength to end-users. To upgrade a conventional TDM-PON to a WDM-PON currently involves replacing the power splitter in the RN with the wavelength multiplexer and replacing all TDM-ONUs at user premises with WDM-ONUs. This all or nothing upgrade is a sort of fork-lifting upgrade that is not only cumbersome but also disruptive to current subscribers and difficult to coordinate. In addition, current WDM wavelength-multiplexers fix the wavelength spacing and optical spectrum at deployment time and constrain future spectral flexibility. In other words, conventional WDM-PON systems use a fixed wavelength plan, which is difficult to change after deployment.
Despite its promise, WDM-PON technologies are still maturing and have not yet achieved mainstream adoption. As such, it is important to have a migration strategy to upgrade TDM-PON to WDM-PON systems seamlessly with minimum disruption to the existing TDM-PON users. Such a system should support coexistence of TDM-PON and WDM-PON architectures during the migration period.